A method and an arrangement of the above kind are disclosed in DE 195 36 038 A1 (U.S. Pat. No. 5,692,472). There, in the context of the control of a drive unit of a motor vehicle, for monitoring purposes, a quantity, which represents an output quantity of the drive unit, is compared to a maximum permissible value, which is pregiven for this quantity. Fault reaction measures are initiated when the quantity exceeds the pregiven permissible value. Examples of the output quantity of the drive unit are: the power of the drive unit or a torque of the drive unit, for example, the indicated torque, the output torque, et cetera. In one embodiment, the computer, which executes the control of the drive unit, includes at least two mutually separate program levels. The described comparison for monitoring purposes is computed in the second program level. The first program level is reserved for programs which compute the functions provided for the control of the drive unit. In another embodiment, a limiting of the input value to the maximum permissible value is undertaken in the first program level. This input value controls the drive unit.
For determining the maximum permissible value, and when no drive command of the driver is present, in general, the largest occurring value of the output quantity, which can be adjusted by the idle control, is inputted. In this way, an unlimited drivability is ensured. Especially in vehicles having small motors, low rolling resistance or a low inner friction, consumers such as a climate control compressor, a torque converter, et cetera, operate greatly on the output quantity of the drive unit so that, with respect to drivability, relatively large permissible values are to be inputted.
To improve the accuracy of the determination of the permissible value of the output quantity, and according to U.S. Pat. No. 6,076,500, an expansion of the maximum permissible value is undertaken for the restart phase for a cold drive unit, whereby, in this region, ancillary functions can operate uninfluenced and simultaneously, outside of this region, a relatively precise determination of the maximum permissible value and therefore a large effectivity is achieved for the fault detection. However, with this method, only two operating states are distinguished.
From the not published German patent application 199 63 759.8 of Dec. 30, 1999, it is known, for determining the permissible value, to compute a weighting with the maximum permissible driver command between a maximum permissible and a minimum permissible value. Here, the permissible requirements of the consumer and of the idle controller are additionally checked via a separate path and considered. With a defective computation of these components, these components are limited.
The described known solutions do not provide optimal results in all cases.